Epicor Kinetic and SAP Ariba integration

What moves between them

The main flow runs from SAP Ariba into Epicor Kinetic. After purchase orders are approved in Ariba, ml-connector reads Ariba purchase orders, invoices, and supplier records on a schedule aligned with your procurement cycle. Purchase order lines are mapped to Epicor vendors and GL accounts, so they land on the correct cost center and vendor account. Supplier data flows both directions to keep vendor masters aligned. Invoice records sync from Ariba into Epicor with approval status, reducing manual entry in accounts payable.

How ml-connector handles it

ml-connector stores both credential sets encrypted and manages OAuth2 refresh on both sides: Epicor Bearer tokens expire in approximately one hour, and Ariba tokens expire in exactly 3,600 seconds. For Ariba reporting jobs, ml-connector submits jobs with the updatedDateFrom and updatedDateTo query parameters within the one-year window constraint, tracks the jobId, polls for completion, and downloads paginated results in chunks. On the Epicor side, it queries purchase orders and invoices filtered by UpdatedOn to find records changed since the last run. Purchase order lines are mapped to matching vendors and GL accounts in Epicor before posting, and supplier records are matched by business entity code. Because Ariba's REST API is read-mostly, ml-connector does not attempt two-way sync on invoices; instead it uses the reporting job pattern to read invoice status and correlate it with Epicor invoice records by external reference number. If a mapping fails, the record is logged in the audit trail and can be replayed once the mapping is corrected.

A real-world example

A mid-sized discrete manufacturer runs Epicor Kinetic for production and procurement finance, and uses SAP Ariba for strategic sourcing and supplier invoice management across their supply base. Before the integration, the procurement team exported Ariba purchase orders every week and manually re-entered them into Epicor, then spent days in month-end matching Ariba-approved invoices against Epicor purchase orders to reconcile the two systems. With Epicor and Ariba connected, new purchase orders sync into Epicor automatically on a weekly schedule, invoices flow from Ariba with their approval status, and the accounts payable team no longer manually re-enters POs or chases invoice discrepancies. Month-end close begins with procurement data already reconciled.

What you can do

• •Sync SAP Ariba purchase orders and invoices into Epicor Kinetic on a procurement calendar, mapped to vendors and GL accounts.
• •Keep Epicor vendors and SAP Ariba suppliers aligned so purchase orders land on the correct vendor accounts.
• •Manage OAuth2 credentials on both sides, refresh tokens before expiry, and handle Ariba's one-year date window constraint on reporting jobs.
• •Track Ariba purchase order approval status and flow it back to Epicor with a complete audit trail.
• •Retry failed records and replay them when mappings are corrected, with full visibility into every sync.

Questions

Which direction does data move between Epicor Kinetic and SAP Ariba?

The main flow is from SAP Ariba into Epicor Kinetic. Purchase orders, invoices, and supplier records move from Ariba into Epicor, mapped to vendors and GL accounts. Supplier master data is aligned in both directions to keep vendor records consistent. Invoice writes use the async reporting job pattern to read Ariba data; ml-connector does not attempt to push invoices back to Ariba's REST API.

How does ml-connector handle Ariba's one-year date window constraint on reporting jobs?

SAP Ariba's async reporting jobs enforce a maximum one-year window between updatedDateFrom and updatedDateTo on a single job submission. ml-connector submits multiple jobs in annual chunks when backfilling multi-year history, tracks the jobId for each, and polls until all jobs complete before downloading paginated results.

What happens if a purchase order cannot be mapped to an Epicor vendor or GL account?

If a mapping fails, ml-connector logs the record in the full audit trail with the reason (vendor not found, GL account inactive, etc.) and marks the record for manual review. Once you correct the mapping in Epicor, ml-connector can replay the record on the next sync cycle without re-querying Ariba.